Spatial Immune Profiling of a Brain Border Region with Multiplex Imaging Proteomics

Recent advancements in spatial biology have positioned it as a rapidly emerging field with significant potential. Imaging proteomics has also seen substantial upgrades, notably in utilizing extended antibody panels exceeding 100 markers. With this innovation, researchers can capture an unprecedented level of biologically relevant data, facilitating the identification, quantification, and profiling of diverse cell populations and tissue structures, as well as mapping their spatial interactions. In this study, we present a comprehensive profiling of a brain border region, the dura mater, using the Akoya PhenoCycler-Fusion platform with a curated 32-antibody panel. Given the complex nature of the data, we specifically developed or refined various computational methodologies to enable a comprehensive analysis. Our approach includes accurate semantic segmentation of diverse immunological cell populations within the dura, utilizing extensively fine-tuned models that sequentially segment multiple input channels. The downstream analysis features (1) morphological profiling of macrophage populations across different conditions, (2) cell-to-cell interaction analysis, (3) vasculature network registration coupled with identification of various types of vessels, and (4) neighborhood analysis. Taken together, these methods provide a holistic understanding of the immune landscape within the brain border region and the relationships between different cellular and structural elements. Lastly, we highlight the challenges in the field that continue to hinder imaging data analysis.